[RFC] New Interest Rate Framework

[mikey]

Motivation

The current interest rate framework on Vesta was decided mainly to fight off significant VST de-pegs. However, the base level of the interest rates (when VST isn’t suffering from de-peg) was arbitrarily decided. Looking across the interest rate charged on collaterals in similar platforms, Vesta’s current interest rates significantly undercut those of competitors, but we are not seeing plenty of deposits to take advantage of these rates.

Specification

Change the interest rate framework to account for both significant de-peg and better value capture by having the interest rate be influenced by both competitors’ rates and the peg. In the future, relay the fee to further deeper liquidity of VST.

Specifically, each asset will have a base rate, which is determined by the competitors’ rates. In the event of a depeg, the interest rate would react accordingly using the currently implemented reference rate technique. For a graphical example, please view the chart below. The x-axis represents the depeg amount in %, and the y-axis represents the total annual interest rate in %.

Base Rate

Base rate is the rate charged when the stablecoin is on-peg. Being on-peg is defined as within the peg threshold. To start, we are defining the peg threshold at ±0.5%. This means that if the token is between 0.995-1.005, then the interest rate should remain at the base rate level.

Different from our previous interest rate approach where the rate was determined by the overall soundness of the collateral, the rate is now determined by looking at competitors’ rates.

How rates are determined for different protocols

• For CDP protocol: we’ll refer to another CDP protocol’s interest rate. If the CDP protocol has multiple risk parameters, we’ll refer to the one with the closest risk parameters.
• Lending protocol: collateralizing against your collateral on a stablecoin protocol is similar to depositing the collateral into a lending protocol and then borrowing stablecoins against it. So we will take that rate.

What if the collateral is not listed anywhere else?
We plan to charge a flat x%, we welcome all suggestions on the specific numbers.

Collateral	Proposed Rate	References	Notes
ETH	0.5%	Maker ETH-B, Liquity
wstETH	2.5%	Maker WSTETH-A, Aave mainnet, Aave Arbitrum, Radiant
ARB	10%	Radiant, Sentiment
GMX	x%	-
DPX	x%	-
GLP	7%	Abracadabra	We plan to lower our fee on GLP’s native APR in this fee change as well
OHM	10%	FraxLend

Special note on GLP

Why are the rates higher than what it is currently?
Currently, almost all of Vesta’s liquidity is funded by an OTC deal using Vesta’s governance token VSTA, and this deal will not last forever. We intend for the protocol to become a public good and be fully self-sustaining, which means that the interest charged will not be going to the core contributors but rather fully going to liquidity provisioning.

Reference Rate

The current interest rate mechanism of reference rate remains. To learn more, please visit Vesta Reference Rate - Vesta.

We plan to abolish the original tier approach and instead use a universal reference rate with some exceptions since there are certain types of collateral that make for better collateral than other ones. This would allow us to attract more of the said collateral, resulting in the stablecoin being backed by better collaterals.

Proposed max reference rates:

Collateral	%
ETH	5%
wstETH	100%
ARB	200%
GMX	200%
DPX	200%
GLP	200%
OHM	200%

Implementation and Future Plan

We plan to proceed with a manual, governance driven approach to start the fee setting process. In the future, we plan to move to an automated process, where interest fee across all competitive products would be aggregated and fed on-chain via oracles.

In the future, we also aim to relay the fee in a way that could further deeper liquidity of VST. This is a subject that’s under active development.

Timeline

This poll/discussion period will be live for a week and if passed, the official voting will commence immediately after and will take place over three days.

[florian]

In favor.

This proposal also adds an important feature for investors: predictability. Previous model was too volatile , this one looks more stable when VST is at peg!

However, I would set 10% for GMX and DPX, 15% is quite heavy.

[narok]

Completely disagree with raising of interest rates.

We are having a product with a lot of competitors and this will be shooting into our own foot for negligible gain at this TVL.

We need to adress product stability by implementing VST use cases and improving liquidity instead of raising fees.

[tylerpol]

This is absolutely moronic. If GMX isn’t listed anywhere else, and I have to pay a 15% APR to collateralize it, I’m better off not doing so at all. You’re going to lose revenue by doing this. Have you not considered the possibility that your vaults aren’t maxed out because the demand simply isn’t there? In other words, the only reason why people are actually using these vaults is because the interest is so low? Demand elasticity is a real thing and if I need to pay an insane APR to take a loan against my GMX, I’m better off just staking it and collecting the multiplier points.

It’s just bizarre logic. You’re basically saying - “Our vaults are underutilized because not enough people are using them. Let’s solve this by removing the only incentive people currently have to use them. Yeah, that’ll do the trick.”

[Maswasnos]

Seriously?

To be blunt, these interest rates are extreme. If the team feels that the protocol is “underpaying” itself and charging too low of a rate for certain assets, fine, but that’s not a reason to 10x the rates. I’d consider even as much as a 2x on the ARB/DPX/gOHM/GLP vault rates given the high adoption of those, but a ~10x is nuts. I’m voting no if the proposal values are anywhere near this high.

I’ve already decided not to put my GMX into Vesta due to the current rates and fees, this would drive away any residual demand for that vault. Why would I use Vesta with my ARB at 10% interest when I can go to Silo.finance and borrow USDC at just under 2% right now? It’s not remotely competitive. I’m not sure how you arrived at 2.5% for wstETH either, that’s significantly higher than Maker’s rates and I think it’s a mistake to compare borrowing VST with borrowing USDC/USDT on Aave. Like it or not, VST simply lacks the utility of USDC/USDT and people aren’t going to pay a premium to use it versus those options.

I don’t mind the idea of abolishing the tier approach given that each collateral asset is different; that seems very fair to me. However, I would encourage the team to treat ETH LSTs like wstETH (or rETH whenever you add that ) more like ETH. If you want to have a higher rate and/or a higher collateral ratio to account for smart contract risk, OK, but I really think we should keep it to a minimal premium. Those kinds of tokens will eventually replace ETH in many protocols and it would benefit Vesta to get ahead of that curve.

If you ask me, the vaults as they currently stand look fairly healthy. We’re seeing high utilization of the ARB, DPX, GLP, and gOHM vaults, and the wstETH vault also isn’t looking too bad. The GMX vault seems to have low overall utilization and I think we might want to consider lowering the APR fee, but it’s otherwise doing okay. The ETH vault is doing relatively well, and I think lowering the APR a bit further to .5% would help attract more capital.

What Vesta really needs is to pursue protocol partnerships and build demand for VST, as Narok said.

[mikey]

Hey all, thanks so much for your replies. It seems like the proposed interest rate numbers got a bunch of attention.

The part that’s missing from this proposal and one I definitely should have included is that the revenue would all be going to fund liquidity provisioning. The problem with this is that it would expand the engineering scope quite a bit, where we need to add the VSTA staking module, and the liquidation flow. In this case, we’ll batch everything into the new, bigger upgrade in a couple months.

I do want to emphasize that such an interest increase is inevitable as current liquidity is funded by Tetranode, which does not last forever. We intend for the protocol to become a public good and be fully self-sustaining, which means that the interest charged will not going to the core contributors but rather fully going to liquidity provisioning.

[Maswasnos]

How much revenue does the protocol need to become independent of Tetra’s contributions? Were those rates in the OP decided with a particular revenue # in mind?

Like I said, I don’t mind increasing rates at this time, it’s just the severity of the rate increase I find disagreeable. The rate for ARB still seems very uncompetitive when protocols like Qi DAO are only charging 2.5% interest on ARB vaults for minting MAI. The GLP vault will also become fairly unattractive at 7% interest since it’s only yielding about 10% APR right now.

Could we just start out by doubling the base APR on ARB, GMX, DPX, GLP, and gOHM and see how that affects protocol usage?

[tylerpol]

Could you share the cost/benefit analysis you’ve done with this? Surely you must have thoroughly looked at how many users/positions you expect to leave the protocol entirely due to insane interest rates vs how much incremental revenue you’ll collect as a result? I.e. if you lose 80% of existing GMX, ARB, and GLP positions, are you still better off with what will remain? Or are you guys just saying to yourselves “zoinks, we need more liquidity, let’s jack up the cost of using our product”?